Leatherhead Food Research. Food Allergens: From Social Stigma to Sampling. 16th November 2011.
Alex Gazzola reports.

Food is everywhere, Tony Hines, Head of Food Security and Crisis Management Manager at LFR, told delegates – and this makes food allergy a complex social issue.

Two topical manifestations of this include the alarming rise in ‘anti-allergy’ journalism, and in ‘allergy bullying’, both online and in schools, which a quarter of food-allergic young people have experienced.

Hines highlighted articles questioning the seriousness and prevalence of nut allergies, including Salon.com’s The Fear About Peanut Allergies is Nuts, and gave real-life cases of bullying, including a dairy-allergic child having milk squirted in his face by a classmate, and the student charged with assault for smearing peanut butter on the forehead of an allergic peer. 

In response to such cases being raised at the House of Commons by the National Allergy Strategy Group (NASG), the Education Department said it would “try and ensure” schools are equipped to ensure this bullying does not occur. Hines said he was disturbed by this response: what is needed is an automatic suspension for any ‘allergy bully’.

Lynne Regent, CEO of The Anaphylaxis Campaign UK, gave an overview on the burden of food allergy. The quality of life for the 3-4% of child / 1-3% of adult sufferers is effected, and with many having related conditions (eg eczema, rhinitis), there is a significant socio-economic impact on health, food and employment sectors.

This is hard to quantify – as is the impact on individuals and families, which includes time demands (seeking healthcare advice, longer periods shopping / information-seeking), financial costs (travel to allergy centres, cost of safe food / medication) and education / employment consequences (lost school days, chances of promotion reduced).

Dr Richard Pumphrey, honorary consultant immunologist at Central Manchester and Manchester Children’s University Hospitals spoke on food allergen avoidance failure. In an often sad presentation we learned of many young fatalities. These included a 13-year-old girl, who’d consumed chicken tikka several times at a certain restaurant, and at a later date ordered the same meal from a different outlet without checking ingredients. It contained 10g of peanuts. Although stabilised one hour after treatment in hospital, a biphasic reaction caused sudden collapse and death in four minutes. Pumphrey speculated such high-dose cases “may not be rescuable”.

Another incident, provoked by thoughtless food labelling, was sobering. “May contain traces of nut” was printed on the product, which lulled a consumer into a false sense of safety, having previously ‘risked’ ‘may contain’ products without problem. Sadly, the ingredients label itself was not consulted. It listed ‘peanuts’. Death came 40 minutes after consumption.

There are, we learned, many such ways to ‘trip up’. There have been cases where poor diagnosis had led to the avoidance of the wrong allergen, and of a forgotten need to avoid an allergen. People are also susceptible to distractions – for instance, in one person’s keenness to secure a vegetarian meal, the ‘no nuts’ check was forgotten. A break from routine, such as staying with relatives, may have consequences. Giving in to ‘just this once’ temptation, peer pressure, chefs not appreciating ‘peanut flour’ means ‘nuts’, walnuts lurking in posh breads… the stumbling blocks are endless.

That said, the greatest proportion of severe and fatal reactions occur when individuals do not expect to find an allergen as an intentional ingredient in a catered food.

Food Adviser to the Anaphylaxis Campaign, Hazel Gowland, spoke about consumer studies into allergen avoidance, drawing from a programme of research from Brunel, Surrey, and Southampton universities for which she was adviser. Its aims were to understand food allergic individuals’ ‘rules of thumb’ when choosing foods, and their approach to risk assessment and labelling information – all undertaken via accompanied shops, interviews and reasoning tasks.

The voluntary status of allergen advice boxes is not widely known, said Gowland, and its absence was considered by many to imply an allergen-free product. Ingredients lists are considered useful, especially for new products, but not always easily readable. Inferences are drawn from product shots (eg a smooth looking food is taken to mean no nuts) and from a product’s name (eg a word like ‘mix’ rings alarm bells).

Overwhelmingly, allergic consumers feel ‘may contain labelling’ is, as one put it, a “backside-covering exercise” by manufacturers, and many run the risk of eating such products, for instance when they assume any possible reaction may be minor or when the brand is trusted.

New labelling regulations starting to come into effect next year are a positive step: the 14 key allergens will need to be highlighted on ingredients lists, either through bolding or alternative colour / font.

Antony Bagshaw of Leatherhead Food Research’s Food Chemistry Laboratory discussed allergen testing and its accuracy. The three approaches are:

* ELISA (Enzyme Linked Immunosorbent Assay) – the most common method, which detects protein molecules.
* PCR (Polymerase Chain Reaction) – which detects DNA, and tends to be used when ELISA kits are not available for specific allergens.
* LCMS (Liquid Chromatography / Mass Spectrometry) – precise, but currently prohibitively expensive.

Using ELISA, Bagshaw demonstrated the limits of measurability – and the associated uncertainty.

ELISA measures an immune response, with antibodies on a plate capable of detecting an allergenic protein. The sample itself, applied to this plate, is the biggest problem: how accurately does it represent a larger sample of the food product? There is often a difficulty in extracting the protein or allergen, and the fact that the test may measure not the protein that individual(s) react to, but one which has been designated representative of the whole food, introduces another element of uncertainty.

Furthermore, different test kits produce different results. In one set of examples provided of peanut detection in chocolate, figures varied enormously, from 4ppm to 54ppm. But this is, Bagshaw said, “as good as it gets”. We can accurately determine presence or non-presence of an allergen, but quantification remains tough. Where with gluten a measurement of, say, 10ppm is safely within the gluten-free threshold of 20ppm maximum, with most other allergens, total absence is ideal. Yet this presents the problem of detection limits: how low can ELISA go? Even if as low as 2.5ppm, there are exquisitely sensitive individuals who may react to 1ppm…

Dr Joanna Topping, senior scientist at LFR, voiced the shaky assumption of this limitation: that we are effectively and for convenience saying that the limits of what we can detect is the safe limit – and yet there is no basis to believe this might be true.

She explored other difficulties. There are many allergenic proteins in egg, for instance – which one should be measured? White or yolk proteins? Raw or cooked? Food processing changes protein structures, and these may not be detectable. Roasting peanuts at temperatures above 140C leads to sharply declining detection, and different kits produce highly variable results for egg / milk in baked versus unbaked biscuits.

Dr Chun-Han Chan of the Food Standards Agency spoke on progress towards establishing clinical thresholds with a view towards ‘action levels’.

An individual’s threshold falls between the two dosage levels given during an incrementally increasing challenge test at which the latter dose elicits an adverse reaction and the former does not. Allergen management thresholds are levels in foods below which we would not expect significant reactions in people sensitive to that allergen. Their aim, Chan told us, is to protect consumers at the public health level – not to protect every individual on every occasion against any reaction.

The FSA’s Allergen Threshold Programme will work towards this goal: it will draw from EuroPrevall research data, upcoming 2012 studies and input from international bodies and clinicians. A key aim is to establish levels suitable for ‘not suitable for’ and ‘free from’ labelling, which would be consistent, clear on risk, and give meaningful advice – hopefully reducing the flurry of difficult-to-assess warnings we have now (eg ‘made in a factory …’, ‘made on a line…’ and variations thereof).

Dr René Crevel, speaking on behalf of the FDF Allergens Steering Group, wrapped up with a call for a move towards a risk-based approach to allergen management – not the hazard-based approach which presently dominates. (The ‘hazard’ of peanuts somewhere in the factory may translate to a tiny ‘risk’ if strict segregation and allergen management is in place, for example – and yet labelling rarely reflects this. It tends to be dictated by hazard.)

A standardised approach to risk assessment, which is quantitative, will minimise risk and maximise food choices – as well as re-establish trust in labelling among consumers, and consequently reduce risk-taking. Scientifically established action levels will, as Chan said, set “free from” thresholds (defined as “no mild reactions in the vast majority of highly sensitive allergic individuals”) and “not suitable for” thresholds too (“no severe reactions in the vast majority of allergic individuals”). (‘Vast majority’, in this case, would be 95-99%.)

It’s a collaborative effort, stressed Crevel, with food regulators, manufacturers, health professionals and allergic individuals all having a share in the responsibility in reaching these goals.

Greater involvement, education and awareness were recurring themes during the day’s talks – and, despite the complexities of the issues involved, one was left with a feeling that we were on the right track.

Leatherhead Food Research
Food Standards Agency
The Anaphylaxis Campaign
The Food and Drink Federation


Back to top

More conference reports